Torticollis is the primary cause of cranial flattening. How to spot and manage it. Get a coffee as this is a long one…

I was introduced to the clinical condition of plagiocephaly long before I came to the realization that cranial flattening simply would not occur unless there was some factor (almost invariably torticollis) restricting normal head movement in affected infants. The first clue was a significant body of research and personal observation that only a fraction of kids who are slept on their back actually develop clinically relevant flattening. In fact, this percentage has been estimated to be between 15-20%. Stated differently, the vast majority of newborns placed supine after birth do not get flat at all. The next question was to figure out why. As a young researcher, I considered the various theories, such as the at-risk infants had an abnormally soft skulls. Of course, there was no scientific basis for such conjecture and mere observation proves this to be untrue- the infant cranium does not immediately deform when laid on a flat surface. Instead, the process takes time to evolve. Most studies show that the average age parents detect flattening is between 6-8 weeks. This all began to congeal during a trip to the pumpkin patches in New England. As I evaluated the pumpkins, and the various shapes they assumed, it became quite obvious that cranial deformation results from prolonged and sustained external pressure to one area as the infant cranium grows. The pumpkin analogy provides a useful and accurate model for cranial flattening that I still use when talking to parents. I did not appreciate at that time that there was historical support for this theory. Intention cranial deformation had been practiced by many cultures including Ancient Peru, the Chinook Indians of the Pacific Northwest, and the French aristocracy. Each population used sustained pressure over the growing cranium to create permanent changes in skull shape. It became abundantly clear that in order to develop a flat spot on the infant cranium, the infant must lay on that specific area most if not all of the time during the first several month of life.

The inquisitive mind would immediately ask what would compel (or force) an infant to maintain a single head position? Many early clinical reports on plagiocephaly observed that affected infants often demonstrated a head “rotation preference”. But why would some infants, and not others, “favor” a specific head position? Perhaps because of my background and training as an orthopedic surgeon, I began to measure the head rotation of all infants in my practice who presented with plagiocpehaly. I found that nearly every patient had a measurable head rotation discrepancy and that the overwhelming majority of parents reported that the infant turned to one side repeatedly during the first few month of life, often is spite of attempts to reposition the head, the objects in the room, etc. The results of this led to one of my earlier studies on congenital muscular torticollis and plagiocephaly, see https://pubmed.ncbi.nlm.nih.gov/19182625/. It is, in my experience, rare to see deformational cranial flattening without some element of cervical imbalance, cervical tightness, or torticollis.

Any discussion about deformational cranial flattening rings hollow without first acknowledging the elephant in the room: torticollis. This term connotes a “twisted neck” and is essentially a cervical muscle imbalance most commonly from positioning in-utero. Although other smaller cervical muscles can be involved, the primary issue is with the sternocleidomastoid muscles (SCM), the large muscles running front to back along each side of your neck. In the classical presentation, the head is tilted to one side (tight SCM muscle) and rotated to the other side (weaker SCM muscle). There are many theories about the pathogenesis, but most commonly this is from prolonged head immobilization in the pelvis. A good analogy to understanding this condition is casting ones elbow in a bent position- the muscles on one side of the arm become contracted or tight, while the ones on the back of the arm are elongated or stretched. The longer the arm is casted, the worse this imbalance becomes. Cast for 3 weeks, stiffness is minor, Cast for 6 weeks, the stiffness is worse. Cast for 12 weeks, and there may be muscle contracture that is very hard to resolve. This is what happens with torticollis- the longer the head is immobilized in the pelvis, the more persistent the muscle imbalance.

After birth, infants are weak and the imbalance will force the head to one side or position. As the infant develops strength and better head control (about 3-4 months of age), most will stretch out the tighter muscle and acquire the ability to rotate their head to the formerly “tight” side. The weak muscle on the opposite side takes much longer to build tone and this can lead to a persistent head tilt, especially when the baby begins to sit up, long after restoration of normal head rotation. This has led some researchers to suggest that the asymmetrical head flattening actually CAUSED the torticollis. This is totally incorrect as every good clinical study on torticollis shows progressive improvement with growth. A more tenable explanation for this finding is that sitting up unmasks the residual weakness of the muscles since the infant is now supporting their head against the pull of gravity. Makes more sense, right?!! The absence of an identifyiable head tilt until the child begins to sit with assistance also lends to parent and clinician confusion about whether the patient had torticollis at birth. The answer is yes, with rare exceptions.

In some instances, the contraction may persist and require adjuvant treatment such a Botox and/or surgery. Unresolved torticollis can lead to facial scoliosis, a progressive deformity that leads to permanent and severe disfigurement. This takes year to develop as the face matures. I have only seen several cases in my career indicating that it is rare since the vast majority of cases of torticollis are self-limiting and require no aggressive management. Consequently, I reserve any discussion of Botox treatment and/or surgical SCM release until the child is at least 18 months and only if there is a clear head rotation difference. Many clinicians simply do not understand this condition and the sequence of resolution, and respond to a persistent head tilt (weak muscle) with inappropriate managements. There are several reports of failures with Botox and even surgery for this type of tilting- I submit that these authors were treating the wrong muscle! The correct approach would have been continued observation and muscle strengthening of the weak muscle. It is for this reason that I abhor the use of TOT collar or any other orthotic that artificially supports the head. The remedy for a weak muscle is exercise and strengthening, not mobilization and bracing.

Although most authors acknowledge the causal contribution of torticollis, the association between plagicephaly and torticollis is weaker in some reports. I would submit several reasons for the disparity in results. One of the primary reasons clinicians do not recognize torticolllis in infants with plagiocephaly is a fundamental misunderstanding of the underlying condition, its pathogenesis, and its many presentations. For those who deal with this often, it is not so difficult to understand. For those that do not, it can be enigmatic. For example, I ran across a paper by one well- known craniofacial surgeon who routinely operates on patients with torticollis in which he suggested that the diagnosis requires the presence of a palpable mass in the sternocleidomastoid muscles. In fact the mass, inaccurately termed “pseudo tumor”, is an extremely rare finding in patients with torticollis and signals a more severe presentation often associated with intramusclar hemorrhage, fibrosis, or some type of muscular injury. The vast majority of infants with a measurable head rotation difference, or clinical evidence of a cervical muscle imbalance, do not present with this finding. Similarly, some authors diagnose torticollis only in patients with a classic clinical presentation as defined above (head tilted to one side, chin rotated to the other). Although some patients do present in this fashion, many more subtle cases do not. Lastly, some authorities look for a discrepancy in passive head rotation measurements (i.e., an adult examiner manually rotating or tilting the infant head) to define torticollis. This leads to a systematic under-reporting of mild to moderate cervical imbalance since any adult can manual force an infant’s head to one side in cases where the neck muscle is mildly or moderately tight.

As a rule, any infant with a strong head position preference early in life has torticollis until proven otherwise. This group comprises the majority of infants who are at-risk for developing cranial flattening. The most sensitive test for torticollis and/or cervical muscle imbalance is to induce the infant to turn voluntary to each side. Infants who are not able to sit (>3 months) can be assessed by laying them flat, and stimulating them to turn to either side by rattling keys, crumpling paper, or a parents voice. Those who can sit should be induced to look to either shoulder (must stabilize the shoulder to avoid torso rotation). In most patients, a measurable rotation discrepancy will be identified. The head rotation difference can be due to a tight SCM muscle, a weak contralateral SCM, or in many instances, both. Remember that congenital muscular torticollis is an imbalance of the cervical muscles, and that implies that one sternocleidomastoid muscle is tighter, while the other is weaker (looser). Some infants with resolving torticollis will have normal rotation to the affected side (<20 degree angular measurement- see text), but over-rotate to the weak side (able to touch chin to shoulder). This represents a weak SCM on that side and can often manifest as an intermittent head tilt when the infant is fatigued or preoccupied. It is important to recognize that the tight of contracted SCM muscle will tend to loosen up long before the weak SCM of the opposite side will strengthen. Thus, an intermittent head tilt is quite common as torticollis is resolving and should not be confused with a persistent tight SCM on the opposite side (which would manifest as limited head rotation).

Another major reason clinicians fail to identify torticollis in infants with plagiocephaly is that torticollis almost invariable improves with neuromuscular development. Thus, if one is examining an 8 month-old infant with head flattening, the neck imbalance may have already resolved. This should not be interpreted to suggest that torticollis was not present at the time the flattening occurred. If you ask parents of infants with flattening whether the infant has a strong head position preference in life, they will invariable say yes but that the tendency has improved. This was one of the principal findings in our study cited above. Many clinicians categorically refer infants with torticollis (and by association, plagiocephaly) for physical therapy, but often too late to have much effect on cranial shape. It is still important for a PT to evaluate and monitor the progress of the torticollis regardless of when the referral is made, but studies show that PT implemented after 4 months of age is unlikely to alter established head shape significantly. If torticollis is diagnosed within the first 2-3 months, I recommend the Perfect Noggin to avoid further flattening and to improve what has already developed. After 4 months of age, the decision then is to observe or consider a molding helmet for more severe case. That is a discussion for another day!!!